You might have had your suspicions, but now they are confirmed. Many e-mailers do indeed fall into two main categories: the day laborers, who only e-mail at work from 9 to 6, and the e-mailaholics, who constantly check their e-mail during all waking hours of the day.
That’s the result of research done by Dean Malmgren, a graduate student in chemical and biological engineering at the McCormick School of Engineering and Applied Science at Northwestern University, with his adviser, Luis Amaral, associate professor of chemical and biological engineering, and two collaborators from Yahoo! Research. Their paper, “Characterizing Individual Communications Patterns,” will be presented at Knowledge Discovery and Data Mining (KDD), a computer science conference this summer in Paris.
The group studied data from about 3,000 e-mailers at a European university over an 83-day period and 122,000 e-mailers at a U.S. university over a two-year period. The data was anonymized, but researchers could see the time at which each e-mail was sent, and naturally, patterns emerged.
“We wanted to come up with a way to describe how people behave,” says Malmgren, who interned at Yahoo last summer. “Once someone starts doing an activity like e-mail, they are much more likely to continue doing it until they snap out of it and do something else. In this way, people’s activity patterns are very periodic. They have daily and weekly cycles of activity.”
When it comes to e-mail, “day laborers” send e-mail between 9 a.m. and 6 p.m. but not at other times. “E-mailaholics” send e-mails throughout their waking hours — from 9 a.m. to 1 a.m.
“We strongly suspect there are more groups, and that’s a problem we’re currently pursuing,” Malmgren says. But the data provided Malmgren and Amaral with enough information to create a model that describes patterns in which people send e-mail.
“From a sociological point of view, it’s another way to characterize humans, just like gender, race, and network attributes,” Malmgren says. “From a more practical point of view, it might be possible to use this data in real-time to detect spam. For example, if your e-mail account is suddenly sending e-mails all day, even when you’re sleeping, it could be sending out spam. That’s something you can’t otherwise detect with existing methods.”
To view the paper, click here. [Source: McCormick News]